Definition of "Success"
My definition of success is based on:
The potential value of the product that is deliverable to the final customer.
That is, at the point when the product is handed off from the project management team to the project owner. If it is a clearly sufficient product delivered on time and on budget, then the project is a success, regardless of what the rest of the enterprise subsequently does with it. That's important, because once the product is handed off, the project management team has no further control over the business outcome.
However, if that product is incomplete, of poor quality, over budget, or late, then the project has failed. The key element here is that:
Success is measured by the product i.e., the result - and nothing else.
This metric requires that we also do a really good job of scope control, without which our yardstick may move all over the place between inception and completion.
By their very nature, class 1 projects will have, in general, a lower success rate than class 5 projects, simply because they are more difficult, and rely less on established knowledge and practice. This may overshadow both the quality of the team and the process they use. In addition, they exhibit a wider variability in their result as opposed to class 5 projects, whose outcome can be more accurately predicted.
So, in order to begin the analysis, you must first decide which of the five broad classes most resembles your project. The examples provided from different domains should help you distinguish the extremes from the middle with two additional halfway points for somewhat better granularity. Your project may not be an exact fit, but it certainly should resemble one of the five classes more than the others.